pick_the_cup_demo_dataset

该数据集是一个用于机器人操作任务的数据集，特别关注模仿学习。它包含机器人的状态观察（如末端执行器位置、旋转轴角度、夹持器宽度等）、图像观察（来自摄像头0的图像）、动作数据以及时间戳、帧索引等信息。数据集结构详细描述了每个特征的数据类型、形状和名称，适用于机器人操作任务的研究和开发。

This dataset is designed for robotic manipulation tasks, with a particular focus on imitation learning. It includes robotic state observations (such as end-effector position, rotation axis angles, gripper width, etc.), image observations (images from Camera 0), action data, as well as supplementary information like timestamps and frame indices. The dataset structure elaborately specifies the data type, shape and name of each feature, and is applicable to the research and development of robotic manipulation tasks.